汉语成语识别的认知机制研究

发布时间：2018-05-31 23:30

本文选题：成语 + 心理表象　；参考：《天津师范大学》2014年硕士论文

【摘要】：成语属于非直义语言的一种,以“四字格”为基本形式,是“人们长期以来习用,形式简洁而意思精辟的、定型的词组或短句”,具有结构固定性、语言文言性、意义完整性和语言书面性的特点。影响成语认知的因素有词频、熟悉度、透明度等因素,其中表象性作为影响成语认知的一个因素,一直以来很少得到研究。表象指的是“人们对感知过的事物形象的反映,是事物不在眼前时关于事物的心理复现”,具有直观形象性和概括性两个基本特征,包括视觉表象、听觉表象等。本研究认为在对高表象性的成语进行加工时,同时激活了言语编码系统和表象编码系统,而对低表象性的成语进行加工时,只激活了言语编码系统,因而高表象性成语的加工速度比低表象性成语的加工速度更快,更容易被识别。本研究通过三个实验来考察影响成语加工的认知机制。实验一采用内隐任务,即将成语和非成语混合呈现,让被试对呈现的四字组合进行是否是成语的判断。实验二采用外显任务,让被试对呈现的成语进行表象性判断。实验三结合ERP技术,探究成语认知加工的神经机制。在三种实验条件下,高表象性成语的反应时均比低表象性成语的反应时短。同时,实验三还发现,成语和非成语均引发了N200(170-230ms),且两者的波幅没有显著差异。与成语相比,非成语引发了更负的N400(350-450ms)和更正的LPC((晚正复合体)420-600ms),两者差异显著。与高表象性的成语相比,低表象性的成语引发了增强的N400(350-450ms)和LPC(420-600ms),这些结果可以用双重编码理论来解释,在进行词汇判断任务时,高表象性的成语存在表象编码和言语编码两种编码方式,使得被试能够很快地从低表象性的成语和非成语中区分出来高表象性的成语,而低表象性的成语只有言语编码一种编码方式,使得其比高表象性的成语加工需要更多地认知资源,需要被试投入更多努力,因此导致其识别时间加长,引发增强的N400和LPC。通过实验发现,表象性也是影响成语加工的重要因素。
[Abstract]:Idioms belong to a kind of non-direct meaning language, which takes "four-character case" as the basic form. It is "used by people for a long time, with concise and concise form, stereotyped phrases or short sentences", with fixed structure and classical Chinese. The characteristics of completeness of meaning and written language. The factors influencing idiom cognition include word frequency, familiarity, transparency and so on, among which representativeness, as a factor affecting idiom cognition, has been seldom studied. The representation refers to "the reflection of the image of things perceived by people and the psychological reappearance of things when things are not in sight", which has two basic characteristics of visual visualization and generality, including visual representation, auditory representation and so on. In this study, we found that the speech coding system and the representation coding system were activated when processing the high representation idioms, but only the speech coding system was activated when the low representation idioms were processed. Therefore, the processing speed of high representation idiom is faster than that of low representation idiom. The present study examines the cognitive mechanisms that affect idiom processing through three experiments. In experiment one, implicit task is used, that is, the idiom and non-idiom are mixed, and the subjects are asked to judge whether the four-character combination is an idiom or not. In experiment 2, explicit task was used to make the participants to judge the expression of idioms. Experiment three, combined with ERP technology, explored the neural mechanism of idiom cognitive processing. Under the three experimental conditions, the reaction time of high representation idiom is shorter than that of low representation idiom. At the same time, the experiment 3 also found that both idioms and non-idioms caused N2000-230 Ms, and there was no significant difference in the amplitude between them. Compared with the idiom, the non-idiom elicited a more negative N400 350-450 Ms and a corrected LPCN (late positive complex 420-600 Ms), the difference was significant. Compared with high-representation idioms, low-representational idioms lead to enhanced N400 350-450 Ms) and LPC-420-600 ms.These results can be explained by the theory of double coding when performing lexical judgment tasks. There are two coding methods in high representation idioms: representation coding and speech coding, which enables the subjects to quickly distinguish the high representation idioms from the low representation idioms and non-idioms. But the low representation idiom only encodes a kind of coding method, which makes it need more cognitive resources than the high representation idiom processing, and needs more effort, which leads to the longer recognition time and the enhancement of N400 and LPC. Through experiments, it is found that representativeness is also an important factor affecting idiom processing.
【学位授予单位】：天津师范大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：B842.1

【参考文献】